Hypoglycemia is the most common metabolic disorder of the newborn, and the only known common preventable cause of brain damage in newborn babies. However, the best approach to diagnosis and management remains unclear. One major challenge is that the blood glucose concentration at which brain damage is incurred may differ in different babies, and it is not known how low, how often, or for how long low blood glucose concentrations must occur before brain damage occurs. There have been calls, including from the NICHD1, for large follow-up studies of children in whom glucose concentrations were measured in the newborn period and encompassed a range wide enough to determine relationships between glucose concentrations and later outcome. This proposal is to conduct just such a study. Our hypothesis is that developmental performance in early childhood is related to the severity, duration and frequency of low glucose concentrations in the first week after birth. The specific aims are to determine: 1. Whether the severity, duration and frequency of low glucose concentrations recorded in the first week after birth are related to clinical and developmental outcomes at 2 and 4.5 years of age. 2. Whether these relationships differ in babies with different risk factors or perinatal histories. 3. What glucose concentrations provide the best discriminatory definition for low in relation to these outcomes in different groups of babies, and thus may provide an evidence-based threshold for intervention. A cohort of <500 term and late preterm babies (e35 weeks) is being recruited. All are at risk of neonatal hypoglycemia, and are managed according to current clinical guidelines. All babies also have continuous interstitial glucose monitoring for at least 48 h and up to 7 d after birth. Data from these monitors are not used for clinical management, but do provide valuable information about the severity, duration and frequency of low glucose concentrations. Approximately 40% of the babies develop hypoglycemia, clinically defined as a blood glucose concentration <2.6mM, and are therefore treated. However, a further 20-30% experience episodes of interstitial glucose concentrations <2.6mM that are not detected clinically and therefore are not treated. Children of consenting parents will be assessed at 2 and 4.5 years of age. The assessments will include standardised measures of growth, neurological function, vision, cognitive and language development, memory, executive function, general health and family environment. Data analysis will include multivariate modelling to assess the relationships between neonatal glucose concentrations and developmental outcomes, with appropriate adjustment for perinatal events and potential postnatal family and environmental influences. Results will provide a much-needed evidence base to inform clinical practice in the short term, while also providing the critical evidence base upon which randomized trials of appropriate interventions can be designed.